The long-term objective of this research is to gain an understanding of neural mechanisms and information processing principles involved in sensory acquisition in vertebrate sensory systems. The research is focused on those aspects of sensory processing in which the nervous system actively influences the content and quality of incoming sensory information. In general, the neural mechanisms under investigation fall into two categories, those involved in the control of filtering properties in sensory processing pathways (adaptive signal processing aspects) and those involved in the active positioning of peripheral sensory structures (motor control aspects). The research proposed here will investigate both adaptive signal processing and motor control aspects of sensory acquisition in the electrosensory system of the weakly electric fish, Apteronotus leptorhyncus (brown ghost knife fish). Adaptive signal processing studies will explore the neural mechanisms involved in the descending modulation of stimulus filtering properties in the first-order electrosensory nucleus (the electrosensory lateral line lobe, or ELL). Motor control studies will quantify the behavioral strategy use by these fish when carrying out electrosensory discrimination tasks. The research approach relies heavily on computer modeling and simulation techniques to elucidate underlying neural mechanisms. Biologically-detailed computer simulations of electrosensory processing will be constructed, including: (1) finite-element simulations of peripheral electrical image formation, (2) time-domain models of primary afferent response dynamics, (3) compartmental models of ELL pyramidal cells, and (4) network models of electrosensory processing in the ELL. In vivo single-unit recordings from primary afferents and first-order electrosensory neurons will be carried out to guide and constrain the development of the computer models. These experiments will involve quantifying the threshold, gain, and time-domain response properties of primary afferents and the gain and spatiotemporal tuning properties of pyramidal cells in the ELL. The behavior of weakly electric fish performing electrosensory discrimination tasks will be recorded on videotape and subsequently analyzed to extract information about the control of body position and velocity during sensory acquisition. Spatiotemporal patterns of afferent activation will be reconstructed and compared with the spatiotemporal tuning properties of first-order electrosensory neurons. Stochastic optimal estimation theory will be used to test the hypothesis that the stimulus filtering accomplished by the primary afferents and first-order electrosensory neurons is near-optimal for extracting estimates of the size and location of nearby objects from the electrical images generated on the fish's body surface.